February 11, 2025 ; Catergory: essays
This short essay is essentially a paraphrasing of a conversation that I had not too long ago with a friend. They’ve been working on an emulator as a part of a project for their degree - specifically the PDP-1, which is notable in video game circles for being the machine that the game “Spacewar!” ran on (itself notable for being one of the first examples of what we’d now consider a video game). It’s this factoid that made it a point of fascination for my friend - enough to work on a fully fledged emulator for the system. Hearing about this led me to thinking about emulator development in general (both technical and philosophical), thoughts in the latter which I find to be sufficiently interesting and even novel in places such that I’ve decided I’d note them down here for anyone interested in hearing my yap.
Before continuing I want to at least acknowledge the seemingly never-ending pseudo-philosophical debate on the ethics or legality of emulators, especially as it pertains to the playing of video games or other commercial products. It’s an annoying debate though one I know matters to at least some. Hell, even I at 14 in a (now-private) video discussed it and threw out opinions I certainly do not agree with now. Legality aside (and legality as far as I’m aware is on my side), I’m enough of an anarchist/libertarian for me to want to side with the efforts of preservation and alternate means of access and distributed and free access over the desires and needs of corporations and profit. I’m pro-emulators by default. The rest of this essay will not concern itself really with ethics but this is important groundwork to lay before continuing.
So getting back to my friend’s emulation project - I mentioned that their key interest in reviving the PDP-1 is its connection to video game history. It’s due to this that their first action when building their emulator was to make sure it could properly run Spacewar. But PDP-1 is more than its association with Spacewar - as noted by the Computer History Museum, the PDP-1 was also notable for being the first commercial interactive computer, arguably ushering in an industry wide shift towards ‘minicomputers’. It also inspires the birth of what would become ‘hacker culture’ at MIT. This is where Spacewar was created, but also early advances in computer music. These are arguably also important parts of the legacy of the PDP-1 - or at least are important to the person behind the Computer History Museum article. What we find here are competing visions on what the PDP-1 should be remembered for - different viewpoints on why it matters to people. These viewpoints matter - they matter for what my friend is doing, and they matter for the Computer History Museum, who worked on a restoration project for the PDP-1. It matters because it shapes their efforts. An emulator, as is a restoration, is in conversation with the original machine.
Before expanding on the intricacies of what this conversation looks like or entails in actual emulator design (or how it may inform emulator design decisions), I want to detour slightly into more abstract territory. I’ll start by making the claim that video games are an art form. I will admit this is a controversial claim and I don’t seek in this essay to justify that claim (though I’ll likely make a post doing so in the future so look out for that!). But if we take it that games are art then we might want to ask about what the vectors of artistic expression are in this case. For visual art we have canvases, pages, digital art programs and so on. For film we have CRTs, OLED screens, the cinema screen and so on. The differences in expression vector isn’t just mere curiousity, they physically change the way in which one takes in, perceives and engages with the art itself. The YouTuber Hbomberguy has made a good video making this case with respect to VHS and CRTs for films. So returning to earlier - what are these vectors in gaming? Well clearly it’s the hardware we play games on right? After all its the hardware that allows software to run in the first place - and gaming is just a subset of software. This is almost correct - for there also exists emulators; software that mimics hardware. In this specific instance we might want to say that the vector is the combination of emulator and hardware (ie. an NES emulator running on a PC), but nonetheless an emulator can serve as a medium that augments one experience with a piece of art by changing the context through which it’s perceived. As such, I believe that we can validly consider emulators as having artistic merits worth considering under the context of video games (in the same way that we might analyse the artistic potential of the canvas when used for the creation of visual art) assuming that video games are art (or at a minimum that at least some games meet the bar for art).
Back to the discussion of emulator design specifically. We may want to push back on this idea that I’ve proposed that the creation of an emulator (or a restoration) is inherently a commentary to some extent by author to original. After all we might want to say that the objective goal of an emulator (for which we may validly say an emulator is ‘good’ or ‘bad’) is to capture authenticity. And here authenticity can be measured in a multitude of ways - whether that’s mimicking CPU cycles or getting expected behaviour from games. As such it’s wrong to say that the creation of an emulator is a subjective endeavour. Also, emulators are oftentimes large open-sourced affairs where many hands come together to produce the final emulator. Maybe I could have a point when discussing something made by a singular person but can I still claim subjectivity when it comes to these team efforts? Well to dispel that latter point first, just because multiple hands loosely come together to create a work doesn’t automatically dispel it from being a subjective work - movies after all can be argued to be made similarly. Emulators are ultimately directed affairs - and don’t simply arise spontaneously from collective disjointed efforts. There have to be overarching decisions and directives that dictate whether a certain contribution should be considered for the emulator at large.
Working on that former point, while yes there certainly exist commonly agreed upon standards for emulation - we should note that they firstly don’t cover all the design decisions that need to be made surrounding the creation of an emulator and secondly that even then the question of why we value authenticity hasn’t actually been answered. Regarding the first point, we haven’t actually exhausted all elements of an emulators capacity to reproduce experience through mere pointing at clock cycle accuracy. Sure, this means that all the expected actions are run and therefore that all programs meant to run for the original hardware run correctly, but an experience with a machine is more than just the correct running of the program. There’s how you interfaced with it. The screen it displayed on originally. The input devices it used. Is it more ‘authentic’ to try and display the graphics as cleanly as you can, even though the original could never display graphics that well? Is it more authentic to keep existing memory and CPU limitations hard-coded so that it can’t ever perform better than it could in the day? One may even start to worry if making an authentic machine is even possible; is it not a core part of the ‘authentic experience’ of using the PDP-1 to be there, in 1959 as it came out, with the context of what had come before and to experience its innovations in the early 60s? Is it not inherently an anachronism to take this machine from the past and make it playable today - divorcing it from its original context? Inherently then, the emulator must make decisions when it comes to what counts as authentic to the original machine, and it isn’t clear that there even exists clear answers for what actually is the ‘right’ solution. Hence emulators need to make subjective calls within these domains.
If I were to propose an even more radical thesis, I may even want to say that the choice to favour CPU clock cycle accuracy isn’t strictly an objective choice. That might seem absurd, after all why else do we create emulators? And yes, while it’s true that emulators are defined as something that mimics a form of hardware - whether it does so to a sufficient standard is relative to its use case. So for example, say there existed a computer that we in the present only care for because of its game library. It could do other things outside this but no one is interested in using an emulator of this machine for that purpose. Say then an emulator is created that can run the games well but cannot run the non-games applications well. Is this a bad emulator? For those looking to use the emulator as a games machine, it is a good emulator for them because it suffices for their use case. We might want to say an emulator is incomplete, but it does not follow from this that it has an objective value judgement of being bad, or that it will become an objectively better machine - especially if users needs are being fully met by the existing machine. But if value is instead exported from use case, then its based on the shifting whims of users and consumers - which leads it away from objectivity. It’s just that for most emulators what people want aligns very broadly with CPU cycle accuracy and therefore we can come to mistake the latter for being an objective measure rather than manifesting downstream of the former.
Ok, so let’s take on board this idea that emulators exist in cultural conversation with the hardware for which they are based off. What can we take from this? Well I believe that emulators exist in a very interesting space where an emulator acts as a sort of meta commentary on the legacy of a machine. As mentioned before, there are many design decisions one must make in the creation of the emulator and each tells a story about how the original hardware ought to be remembered. And I emphasise here on remember. Lets circle back to the PDP-1 there at current exist only 3 working models - all at the Computer History Museum. Most people cannot interface with these machines directly, and the CHM does demonstrations of the machine twice a month to show of Spacewar, graphic demonstrations and music (the CHM itself shaping the legacy of the PDP-1 by choosing what factors of its use are of most importance for those in attendance to come away with). An emulator of the PDP-1 is the only way that most people will have to actually interface with the machines directly. As such it’ll have a large impact on how the machine is remembered. If the designer for the emulator were to design it with Spacewar in mind (perhaps as the ‘default’ program) then that would inextricably tie the legacy of the machine to Spacewar, to implicitly say that the value of the machine is that it was the host to the game. And so, the legacy of the PDP-1 warps and shifts in response to what comes after it, what succeeds it in conversation and is able to have the last word. The PDP-1 itself is now a quiet machine. Few originals remain and those who remember using it directly at the time are dwindling in number too. Soon it will only be those who interface anachronistically, inauthentically, through simulacra of the real thing that get to dictate how history remembers this machine. This is the real power that an emulator holds. It defines not only how you interact with the software found on the original system, it defines to a strong extent the public perception of the machine itself.
As a final note, returning to the point about the artistic value of emulators - this leads me to wonder if there could be room for interesting art made by explicitly messing with emulators (or even by building emulators that seek to achieve non-standard behaviour). For example could one maybe heighten a specific emotional response within a game by having the emulator act in non-perfect ways (say to really work on the horror elements of a specific title). Or maybe games can be created specifically with certain emulator features in mind - as to in fact be more authentic to their own visions by explicitly not being on original hardware (and in itself making some form of commentary through this deliberate choice). I’ll admit that I’m really just throwing out ideas here in terms of what seems to be interesting outcomes from viewing emulators from a strictly artistic perspective but I come out of it thinking that there is certainly untapped artistic potential waiting for someone who is willing to experiment with the medium of emulators.